This paper will describe the results of the preliminary phase of a NASA design study for a facility to test high-power electric propulsion systems in space. The results of this design study are intended to provide a firm foundation for a subsequent detailed design and development activities leading to the deployment of a valuable space facility supporting the new vision of space exploration. The objectives for human and robotic exploration of space can be accomplished affordably, safely and effectively with high-power electric propulsion systems. But, as thruster power levels rise to the hundreds of kilowatts and up to megawatts, their testing will pose stringent and expensive demands on existing Earth-based vacuum facilities. These considerations and the access to near-Earth space provided by the International Space Station (ISS) have led to a renewed interest in space testing. The ISS could provide an excellent platform for a space-based test facility with the continuous vacuum conditions of the natural space environment and no chamber walls to modify the open boundary conditions of the propulsion system exhaust. The platform would be designed to accommodate the side-by-side testing of multiple types of electric thrusters currently under development and thus provide a strong basis for comparing their relative performance. The utility of testing on the station is further enhanced by the human presence, enabling close interaction with and modification of the test hardware in a true laboratory environment. These conditions facilitate rapid development and flight certification at potentially lower cost than with conventional Earth-bound facilities. As an added benefit, the propulsive effect of these tests could provide some drag compensation for the station, reducing the re-boost cost for the orbital facility. While it is expected that the ISS will not be capable of generating continuous levels of high power, the utilization of state-of-the-art energy storage media would be sufficient to achieve very high power levels over intervals short enough to be feasible and long enough to provide ample demonstration of steady-state operation. This paper will outline the results of the preliminary phase of the design study with emphasis on the requirements that will dictate the system design.